Extending the UML for Designing Association Rule Mining Models for Data Warehouses

A UML profile for the conceptual modelling of data-mining with time-series in data warehouses

“…Besides the multidimensional model of a data warehouse, we have previously presented conceptual models for commonly-used data-mining techniques such as association rules [10,11], classification [12], or clustering [13]. In all cases, we have used the same technical solution for modelling: profiling UML in order to provide intuitive models which guide analysts through the KDD process.…”

“…1). By following the KDD process, we have proposed a model-based data-mining with association rules [10,11], classification [12], and clustering [13] on top of multidimensional models of data warehouses.…”

A support architecture to MDA contribution for data mining

“…These researches are based on the works of Table 1, so we discuss them at the end of our classification study. The UML profile of Zubcoff and Trujillo (2005) is a UML profile for the association rules technique, but we retain the UML profile of for the same technique, because it is an improved version (Table 1).…”

“…In the literature, we find only the researches of: Zubcoff and Trujillo (2005 and Zubcoff et al ( , 2009Zubcoff et al ( , 2008) that propose to model the data mining process in a very high level of abstraction, by using MDA formalisms, and by considering data mining process only on data warehouses (multidimensional data) (Inmon, 2002). Our work then consists in presenting these researches in the context of the MDA approach, from the defined meta-models, to the implementations, through the transformations.…”